An optoelectronic transducer generally uses a known type of pickup device, such as a PIN diode, which receives the optical signals at the end of an optical transmission line and converts these optical signals into electrical signals that must then be amplified. Many problems arise in designing a transducer of this type capable of effectively processing optical signals when the signals represent numerical data originating in a remote source and when the data rate is high.
Initially, the light signals of the numeric type received by an optical line have several characteristics peculiar to them. One is the very low level of the light signal received, which after being converted by a PIN diode corresponds for example to a current that switches between 0, its value at rest, and a peak value I on the order of 1 .mu.A. The optoelectronic transducer must amplify the signal received greatly so that it can furnish a usable signal having a voltage on the order of 100 mV. If conventional methods are used, the optoelectronic transducer must then include an amplifier having a feedback resistance of approximately 100 K.sup..OMEGA., so as to obtain transfer impedance on the same order of magnitude.
Another particular feature of the light signal received is the very wide frequency band it occupies, which, in conjunction with high output, may be wider than 50 MHz. Given the inherent capacity of the PIN diode, which may be one the order of several picofarads, for the amplifying circuit to be able to pass this band its input impedance must be low, while its voltage gain must be high, at least equal to 200. The present problems in stabilizing the gain and the operating point of the amplifier. On the other hand, the width of the frequency response makes it impossible to use a single resistor of 100 K.sup..OMEGA. as negative feedback in order to define the transfer impedance. The stray capacitances involved will lower its high-frequency impedance.
One known way to stabilize the gain and the operating point of the amplifier is to regulate the mean value of the amplifier output signal to a fixed set-point value by negative feedback. However, this method requires the use of a low-pass filter, which increases the response time of the transducer; that is a disadvantage, particularly when the light signal received is of the intermittent carrier type.